4 ms·

Uh.... Melting USB-C connectors at 65W are already bad enough. The problem is that there is no way to detect a bad contact, and they tend to be. Few specs of

Uh.... Melting USB-C connectors at 65W are already bad enough. The problem is that there is no way to detect a bad contact, and they tend to be. Few specs of dust, and you have 5 amps going to a single pin. Even if you have split seconds momentary disconnects, you can get welds in contact pads, which will over time degrade the contact. On other note, Intel may be increasing laptop CPU power budgets into 60W-70W territory to counter Ryzen people say. I think it makes sense now why they do it.

OnlyOneCannolo 5y ago
Is it really undetectable? A smarter USB PD controller and maybe some extra sensors should be able to mostly avoid that problem, no?
- lmilcin 5y ago
  "smarter" an "should" are key words. Not everybody buys the best hardware in class. Most hardware is cheap Chinese garbage for which the only qualification is that it isn't bad enough to be brought down from Amazon.com. Go explain your grandma or girlfriend why the charger they bought damaged their laptop irreparably.
- nrp 5y ago
  Most devices of meaningful value build in USB-C port protection parts specifically for this reason. Here's a popular one from NXP: https://www.nxp.com/products/power-management/load-switches/nx20p0407-usb-type-c-cc-and-sbu-protection-ic:NX20P0407 https://www.nxp.com/products/power-management/load-switches/...
- baybal2 5y ago
  No, unless you put a sense resistor on every power pin, and add circuitry to individually measure current per pin.
  - zardo 5y ago
    If the alternative is starting house fires that sounds pretty reasonable.
    - Nextgrid 5y ago
      The problem isn't whether reputable manufacturers will do it, it's whether the bottom of the barrel cheap cables from eBay/Amazon will do it. The advantage of USB2 is that it's very hard to screw up. The design is so simple that even the cheapest cable is usualy "okay" because making an "okay" USB2 cable is so simple. In contrast, making a USB-C cable is much more difficult, which means unscrupulous manufacturers flood the market with bad cables that fail with disastrous side-effects.
      - londons_explore 5y ago
        The solution would be for devices to test cables before letting them work. If my iPhone tested the cable was up to spec before charging and said "error, bad cable" if any test failed, then china-cables would be forced to pass all the tests.
        baybal2 5y ago
        If the pin turns bad during use, you can't do much
- megous 5y ago
  Ok, so you detect voltage drop of say 0.4V@5A on the phone end, so slightly out of spec. How do you decide whether it's a slightly underspecced cable (quite common) and 2W being radiated as heat from a 1m long cable (which is not really an issue), or great cable but high contact resistance (where 2W are concentrated into a tiny space that can't cool it)?
  - londons_explore 5y ago
    You could design a port with multiple contacts onto the same cable pin. One contact does voltage sensing while the other takes the current. Then any amount of dirt in the connector can cause whatever heating it likes, but the device can always calculate how much heat is being dissipated in the connector.
    - megous 5y ago
      Nice idea. :) Thanks.
  - baybal2 5y ago
    The cheapest option I see is a fusible link from some very low melting point material.
the_pwner224 5y ago
I used the charger + cable which came with my OnePlus 7 Pro to charge my Samsung Galaxy S8. The cable and port on the phone must have melted and solidified into one unit, because the next morning I couldn't unplug it. With more force the cable came out with the plug damaged and the USB-C male part in the phone ripped in half. I don't think OnePlus makes incredibly high quality & safe chargers like Apple/Samsung, but they're not the cheapest Amazon garbage either. This might be a rare issue, but it does happen. Combined with the mechanical degradation that USB-C ports go through (not as bad as micro-USB, but worse than full size USB-A - A does get loose but still makes good electrical contact), I specifically looked for wireless charging in my next device and try to avoid using the USB port as much as possible.

OnlyOneCannolo 5y ago

Is it really undetectable? A smarter USB PD controller and maybe some extra sensors should be able to mostly avoid that problem, no?

lmilcin 5y ago
"smarter" an "should" are key words. Not everybody buys the best hardware in class. Most hardware is cheap Chinese garbage for which the only qualification is that it isn't bad enough to be brought down from Amazon.com. Go explain your grandma or girlfriend why the charger they bought damaged their laptop irreparably.
nrp 5y ago
Most devices of meaningful value build in USB-C port protection parts specifically for this reason. Here's a popular one from NXP: https://www.nxp.com/products/power-management/load-switches/nx20p0407-usb-type-c-cc-and-sbu-protection-ic:NX20P0407 https://www.nxp.com/products/power-management/load-switches/...
baybal2 5y ago
No, unless you put a sense resistor on every power pin, and add circuitry to individually measure current per pin.
- zardo 5y ago
  If the alternative is starting house fires that sounds pretty reasonable.
  - Nextgrid 5y ago
    The problem isn't whether reputable manufacturers will do it, it's whether the bottom of the barrel cheap cables from eBay/Amazon will do it. The advantage of USB2 is that it's very hard to screw up. The design is so simple that even the cheapest cable is usualy "okay" because making an "okay" USB2 cable is so simple. In contrast, making a USB-C cable is much more difficult, which means unscrupulous manufacturers flood the market with bad cables that fail with disastrous side-effects.
    - londons_explore 5y ago
      The solution would be for devices to test cables before letting them work. If my iPhone tested the cable was up to spec before charging and said "error, bad cable" if any test failed, then china-cables would be forced to pass all the tests.
      - baybal2 5y ago
        If the pin turns bad during use, you can't do much
megous 5y ago
Ok, so you detect voltage drop of say 0.4V@5A on the phone end, so slightly out of spec. How do you decide whether it's a slightly underspecced cable (quite common) and 2W being radiated as heat from a 1m long cable (which is not really an issue), or great cable but high contact resistance (where 2W are concentrated into a tiny space that can't cool it)?
- londons_explore 5y ago
  You could design a port with multiple contacts onto the same cable pin. One contact does voltage sensing while the other takes the current. Then any amount of dirt in the connector can cause whatever heating it likes, but the device can always calculate how much heat is being dissipated in the connector.
  - megous 5y ago
    Nice idea. :) Thanks.
- baybal2 5y ago
  The cheapest option I see is a fusible link from some very low melting point material.

the_pwner224 5y ago

I used the charger + cable which came with my OnePlus 7 Pro to charge my Samsung Galaxy S8. The cable and port on the phone must have melted and solidified into one unit, because the next morning I couldn't unplug it. With more force the cable came out with the plug damaged and the USB-C male part in the phone ripped in half. I don't think OnePlus makes incredibly high quality & safe chargers like Apple/Samsung, but they're not the cheapest Amazon garbage either. This might be a rare issue, but it does happen. Combined with the mechanical degradation that USB-C ports go through (not as bad as micro-USB, but worse than full size USB-A - A does get loose but still makes good electrical contact), I specifically looked for wireless charging in my next device and try to avoid using the USB port as much as possible.